nRF905 on a PCB thinner than standard FR-4 1.6mm

[Witek]

Hello,

We have used the nRF905 in our products over many years. This time we have decided to design the circuit on a 0.8mm FR-4. The recommended layout and BOM suggest to use C11 (please check the "nRF905 RF and antenna layout" nAN900-04).

How do I choose the value for this capacitor?

Thanks

Witek

[Witek]

Hi Jonathan,

Then, I guess, I just look at the signal strenth out of the antenna, to see what value works best?

Thanks

Witek.

[JONATHAN LL]

You could use a VNA to look at the 50 Ohm RF IO line where a antenna would be connected. You can then see how it effects the impedance and if the singel ended match is 50 Ohms. But do you need to use the C11 ? The result should be approximate 50 Ohms without C11. 

Regards,
Jonathan

[Witek]

Struggling with the range at this stage, so I am looking at all possibilities.

If I undersdand you correctly C11 should not make much difference?

[JONATHAN LL]

Are you able to measure the impedance of the line with a VNA ? There will be some change depending on what value that would would put on C11 but to what extent is difficult to say. Have you followed the reference layout closely? Have you tested the antenna before and know that it preforms well and has the antenna been tuned tot he device its on?

Regards,
Jonathan

[Witek]

Currently I dont have access to a VNA, but will try to get one and see. 

We changed the layout silghtly to make it fit on the little PCB it goes on. But I am tihnking I should try to lay out the PCB again, with the RF part exactly the same as the reference.

The antenna is a little spring coil, which is meant to be used in 433MHz range, which is soldered directly on the PCB. I added a matching circuit and matched it well, but the range is still very poor.

This device is a coin cell battery powered sensor, which will transmit only (except for configuration).

[JONATHAN LL]

The layout is sensitive to changes so this will effect what component selection to some degree. Difficult to say exactly what values should be modified, might be fast to just test different values and see if you see any improvements or degradation.

there are some "cheap" pocket VNA's that are starting to get quite good. This might be something to test with as it does preform ok like the LiteVNA and NanoVNA2. 

Regards,
Jonathan

[Witek]

Thanks for the tips. I have redesigned the PCB (waiting for prototypes) to follow the reference design a lot more closely. But will also be looking at the VNA's you suggested.

We have a RigExpert AA-600 antenna analyser. It is not a VNA but, if my understanding is correct, in this case we only are taking S11 measurements. So it should do the job?

[JONATHAN LL]

Yes, that should do the job. 

Regards,
Jonathan